Nonparametric inference for censored data using deep neural networks
Open Access
- 2 April 2026
- journal article
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 88 (5) , 1535-1555
Abstract
We propose a novel deep learning approach to nonparametric statistical inference for the conditional hazard function of survival time with right-censored data. We use a deep neural network (DNN) to approximate the logarithm of a conditional hazard function given covariates and obtain a DNN likelihood-based estimator of the conditional hazard function. Such an estimation approach enhances model flexibility and hence relaxes structural and functional assumptions on conditional hazard or survival functions. We establish the nonasymptotic error bound and functional asymptotic normality of the proposed estimator. Subsequently, we develop new one-sample tests for goodness-of-fit evaluation and two-sample tests for treatment comparison. Notably, we design a new test specifically tailored for testing nonparametric Cox models. The consistency of these tests is established by analyzing the power functions. Both simulation studies and real application analysis show superior performances of the proposed estimators and tests in comparison with existing methods.Keywords
Funding Information
- National Natural Science Foundation of China (12401366)
- National Natural Science Foundation of China (12501387)
- National Natural Science Foundation of China (12271459)
- Research Grant Council of Hong Kong (15307824)
- Research Grant Council of Hong Kong (15306521)
- City University of Hong Kong (7200765)
- City University of Hong Kong (7020159)
- Hong Kong Polytechnic University (P0045385)
- Hong Kong Polytechnic University (CMFAI-P0046811)
- Hong Kong Polytechnic University (JLab-P0052683)
This publication has 42 references indexed in Scilit:
- Genome‐wide association study‐based deep learning for survival predictionStatistics in Medicine, 2020
- Stop explaining black box machine learning models for high stakes decisions and use interpretable models insteadNature Machine Intelligence, 2019
- An Integrated TCGA Pan-Cancer Clinical Data Resource to Drive High-Quality Survival Outcome AnalyticsCell, 2018
- Predicting cancer outcomes from histology and genomics using convolutional networksProceedings of the National Academy of Sciences, 2018
- Mastering the game of Go with deep neural networks and tree searchNature, 2016
- ImageNet Large Scale Visual Recognition ChallengeInternational Journal of Computer Vision, 2015
- Reinforcement Learning and Markov Decision Processes2012
- Applied Survival Analysis2008
- Survival Analysis2003
- A Simplified Method of Calculating an Overall Goodness-of-Fit Test for the Cox Proportional Hazards ModelLifetime Data Analysis, 1998